Is converged infrastructure the answer to today’s data centre challenges?

Steve Phillips, Senior Vice President and Chief Information Officer
Avnet, Inc

In recent blog posts, I’ve discussed a few of the challenges facing today’s data centres, including the growth of data and data analytics, the rise of cloud computing and the growing importance of IT  across all aspects of the enterprise.

While software-defined architectures show a lot of potential to address the flexibility, agility and cost pressures on IT operations, widespread adoption is likely several years away, particularly for established data centres with significant legacy investments.

So what can IT teams do today to help make their current on-premise infrastructure more streamlined, agile and cost effective without fully committing to a next-generation software-defined environment? For many of us, the answer lies in converged infrastructures.

WHAT IS A CONVERGED INFRASTRUCTURE?

The traditional data centre is made up of silos, with each application running on its own server, storage and networking stack, often from different vendors. Provisioning, deploying and managing the hardware needs for each new application requires a significant amount of resources, reducing responsiveness and efficiency and increasing complexity and cost.

While virtualization has helped to increase efficiency and utilisation rates and reduce costs over the last decade or so, in recent years the gains simply haven’t been able to outpace growing IT-related cost and performance pressures.

In a converged infrastructure solution — sometimes referred to as an integrated system —the computing, storage and networking systems from different manufacturers are all brought together by a vendor into a single solution. All the individual components are engineered and pre-configured to work together and with your specific data centre environment, right out of the box.

By delivering a single, pre-configured solution—in some cases, complete with your specific applications already installed—IT teams are able to streamline the procurement, deployment and operation of new hardware, reducing time and cost in the process.

These converged resources arrive on site fully virtualised, pooled and able to be managed through a single interface. This approach increases utilisation rates beyond what virtualization alone can deliver, while also greatly reducing complexity and ongoing maintenance costs. 

CONVERGED INFRASTRUCTURE MARKET LEADERS

The IT market has responded favourably to converged/integrated architectures, according to IDC analysts: “An expanding number of organisations around the world are turning to integrated systems as a way to address longstanding and difficult infrastructure challenges,” said Eric Sheppard, IDC’s Research Director, Storage “[which] makes these solutions an increasingly rare source of high growth within the broader infrastructure marketplace.”

A good example of converged infrastructure at work can be found at VCE, which began as a joint partnership between EMC storage, Cisco networking, and VMware computing and virtualization software.

While Gartner rates VCE, Cisco-NetApp and Oracle as leaders in converged infrastructure, there is healthy competition in a market that IDC forecasts is growing at nearly 20% a year.[1]

Source: Gartner Magic Quadrant for Integrated Systems, 2014

THE BIRTH OF HYPERCONVERGENCE

With the converged infrastructure model gaining real traction, manufacturers are beginning to take the concept one step further with what are being called “hyperconverged architectures.”

Instead of bringing together hardware from multiple vendors, hyperconverged solutions rely on the latest hardware from a single vendor, combining super-dense computing cores, all-flash storage and software-based networking into a single pooled appliance.

These next-generation infrastructure approaches are integrated so tightly in the hyperconverged appliance that they have the ability to deliver computing densities, performance and benefits that aren’t possible in traditional architectures, including:

  • Reduced IT complexity
  • Increased utilisation rates
  • Increased agility and responsiveness
  • Lower power consumption and space needs
  • Lower total cost of ownership
  • Lower procurement, deployment and management costs
  • Increased application performance
  • Lower maintenance costs

A NEW BREED OF DATA CENTRE BRANDS

 

While the convergence market is populated with many familiar IT brands, the more disruptive hyperconvergence model has been embraced by startups and emerging brands like Nutanix and SimpliVity.

This new wave of vendors is looking to disrupt the market, and their innovative model is riding a wave of interest from the venture capital community: SimpliVity has raised more than $275 million in funding to date, with Nutanix attracting an additional $312 million and a long-term partnership with Dell in 2014.

DOES A CONVERGED/HYPERCONVERGED INFRASTRUCTURE MAKE SENSE FOR YOU?

Converged and hyperconverged solutions are a solution worth considering, if you’re part of an established IT shop that needs to:

  • reduce or contain IT expense for on-premise environments
  • respond to the changing needs of your business
  • evolve with the changing trends in IT

In many respects, converged infrastructures are an evolution of virtualization, delivering utilisation and efficiency improvements without having to “rip and replace” legacy investments in the process.

Until the software-defined data centre becomes a reality, converged/hyperconverged infrastructure solutions represent an excellent opportunity to improve the flexibility and expense profile of a data centre without the need for wholesale abandonment of existing data centre investments.



[1] Source: IDC, Worldwide Integrated Systems 2014–2018 Forecast: State of the Market and Outlook

Posted under Converged infrastructure